p38 Mitogen-activated Protein Kinase Is Involved in Fas Ligand Expression*

p38 mitogen-activated protein kinase (MAPK) is activated by T cell receptor engagement. Here we showed that T cell receptor activated p38α but not p38δ. Inhibition of p38α by the specific inhibitor SB 203580 prevented activation-induced cell death in T cells. SB 203580 had no effect on Fas-initiated apoptosis. Instead, SB 203580 preferentially inhibited activation-induced Fas ligand (FasL) expression. The inhibition on FasL expression by SB 203580 was correlated with the suppression on the FasL promoter activation. Overexpression of active MAPK kinase 3b, the activator of p38 MAPK, led to activation of FasL promoter and induction of FasL transcripts in T cells. Stress stimulation of T cells by anisomycin also induced FasL expression in a p38 MAPK-dependent manner. The induction of FasL expression in nonlymphoid cells such as 293T also required activation of p38 MAPK. Our results suggest that p38 MAPK is essential for FasL expression.

p38 mitogen-activated protein kinase (MAPK) is activated by T cell receptor engagement. Here we showed that T cell receptor activated p38␣ but not p38␦. Inhibition of p38␣ by the specific inhibitor SB 203580 prevented activation-induced cell death in T cells. SB 203580 had no effect on Fas-initiated apoptosis. Instead,

SB 203580 preferentially inhibited activation-induced Fas ligand (FasL) expression. The inhibition on FasL expression by SB 203580 was correlated with the suppression on the FasL promoter activation. Overexpression of active MAPK kinase 3b, the activator of p38 MAPK, led to activation of FasL promoter and induction of FasL transcripts in T cells. Stress stimulation of T cells by anisomycin also induced FasL expression in a p38 MAPK-dependent manner. The induction of FasL expression in nonlymphoid cells such as 293T also required activation of p38 MAPK. Our results suggest that p38 MAPK is essential for FasL expression.
Fas (APO-1, CD95) is a 45-kDa membrane protein that triggers apoptosis when it interacts with Fas ligand (FasL) 1 (for review see Ref. 1). The expression of Fas is low in resting T lymphocytes, whereas the expression of FasL is absent. T cell receptor engagement leads to increased expression of Fas and FasL. The subsequent Fas-FasL interaction is the major mechanism underlying activation-induced cell death of immature T cells (2)(3)(4)(5). Fas and FasL are also up-regulated by various stress stimulation. Treatments with anisomycin, UV, ␥-irradiation, or cytotoxic drugs induce the expression of Fas and FasL in T cells and tumor cells (6 -12). Fas and FasL gene promoters have been extensively studied (12)(13)(14)(15)(16)(17)(18)(19)(20)(21). Transcription elements including NF-AT, NF-B, and AP-1 are identified on the FasL promoter (12,13,(15)(16)(17)(18)(19)(20)(21).
In lymphocytes, p38 MAPK is stimulated by stimuli other than stresses. p38 MAPK is constitutively activated in freshly isolated thymocytes (43). p38 MAPK is activated in response to T or B cell antigen receptors and to IL-2 and IL-7 in lymphocytes (44 -48). p38 MAPK is also shown to be activated in T helper 1 cells but not in T helper 2 cells when stimulated by TPA/ionomycin (49). In this study, we examined the role of p38 MAPK in activation-induced lymphocyte death. We observed that suppression of p38␣ by SB 203580 substantially prevented activation-induced cell death in T cells. The inhibition of activation-induced cell death by SB 203580 was attributed to a suppression of the FasL expression. The role of p38 MAPK was further demonstrated by the fact that activation of p38 MAPK by MKK3b increased FasL expression. Our results suggest the possibility that apoptosis induction may be enhanced by p38 MAPK activation through increased expression of FasL.

EXPERIMENTAL PROCEDURES
Reagents and Cell Lines-Concanavalin A, TPA, and A23187 were purchased from Sigma. SB 203580 was a gift of Dr. John C. Lee (SmithKline Beecham, King of Prussia, PA) and was subsequently purchased from Calbiochem (San Diego, CA). The active mutants of MKK3b (MKK3b(Glu 189 ,Glu 193 )) and MKK6b (MKK6b(Glu 207 ,Glu 211 )) were previously reported (30). CAT reporters containing AP-1 and NF-AT elements from the IL-2 promoter were previously described (50). kB-TATA-CAT containing two copies of the HIV B site (51) was a gift of Dr. Warren C. Greene (University of California, San Francisco, CA). Human FasL promoter (Ϫ453 to Ϫ2 nucleotide) was isolated by PCR according to the method of Holtz-Heppelmann et al. (18) and was subcloned into the HindIII and XhoI sites of the pGL2-Basic luciferase reporter vector (Promega, Madison, WI) (abbreviated as pGL2-FasL). Fluorescein isothiocyanate-conjugated anti-mouse Fas antibody Jo2, and biotin-conjugated anti-mouse FasL antibody Kay-10 were obtained from PharMingen (San Diego, CA). T cell hybridomas 10I and 9C12.7, specific for repressor, have been previously used as model cells to study activation-induced cell death (52). Splenic T lymphocytes from BALB/c mice were isolated by panning twice on plates precoated with goat anti-mouse Ig antibody (Sigma) (53). For study on activationinduced apoptosis, splenic T cells were activated with TPA/A23187 for 24 h. Activated splenic T cells were then washed and incubated in the presence of IL-2 (10 units/ml) for another 3 days before anti-CD3 treatment.
Transfection-1.6 ϫ 10 7 T cells were washed once with STBS (25 mM Tris⅐HCl, pH 7.4, 137 mM NaCl, 5 mM KCl, 0.6 mM Na 2 HPO 4 , 0.7 mM CaCl 2 , 0.5 mM MgCl 2 ) and incubated with DNA in 1.2 ml of STBS containing 0.5 mg/ml DEAE-dextran for 20 min at room temperature. T cells were then treated with 15% dimethyl sulfoxide for 3 min and washed once with STBS (54). For luciferase activity, the production of light through oxidation of luciferin in the presence of ATP was measured using a luminometer. For transfections with pGL2-FasL and PGL2-basic, 1 g of pCH110 (Amersham Pharmacia Biotech) was included (50). The luciferase activity was normalized against the ␤-galactosidase activity determined in each transfection.
Quantitation of Fas and FasL mRNA-2 g of total RNA was used for cDNA synthesis by using oligo(dT) as primer. One-tenth of the cDNA synthesized was then amplified by using the following primers: mouse Fas 5Ј-ATC CGA GCT CTG AGG AGG CGG GTT CAT GAA AC; mouse Fas 3Ј-GGT TCT AGA TTC AGG GTC ATC CTG; mouse FasL 5Ј-CAG CTC TTC CAC CTG CAG AAG G; mouse FasL 3Ј-AGA TTC CTC AAA ATT GAT CAG AGA GAG (55); human Fas 5Ј-TGC CCA AGT GAC TGA CAT CAA C; human Fas 3Ј-AAG AAG AAG ACA AAG CCA CCC C; human FasL 5Ј-CAG CTC TTC CAC CTA CAG AAG G; and human FasL 3Ј-CAT TGA TCA CAA GCC CAC C.
Cell Death Measurement-All cultures were performed in RPMI with 10% fetal calf serum (both from Life Technologies, Inc.), 10 mM glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, and 2 ϫ 10 Ϫ5 M 2-mercaptoethanol. The extent of apoptosis was determined by propidium iodide staining. At the end of different treatments, cells were resuspended in hypotonic fluorochrome solution (50 g/ml propidium iodide, 0.1% sodium citrate, 0.1% Triton X-100) (56) and placed at 4°C in the dark overnight. DNA contents were analyzed by FACScan (Becton Dickinson, Mountain View, CA). Fraction of cells with sub-G 1 DNA content was assessed using the CELLFIT software program (Becton Dickinson) (57).
Protein Kinase Assay-T cells were treated with anti-CD3, TPA/ A23187, anisomycin, or hydrogen peroxide in the absence or presence of SB 203580 (10 M). Cell lysates were prepared 20 min after activation, and 100 -200 g of lysate was precipitated with 1 g of anti-p38␣ or anti-p38␦ antibodies (36) for p38 assay, 1 g of anti-ERK2 C-14 antibody (Santa Cruz Biotech) for ERK assay (53), or 1 l of anti-JNK1 Ab101 (58) for JNK assay, followed by 20 l of protein A-Sepharose. The kinase activity of the immune complexes was determined by using GST-ATF-2 (1-109) as substrates for p38 assay, myelin basic protein as substrates for ERK assay, or GST-c-Jun (1-79) as substrates for JNK assay. The reaction mixtures were resolved on SDS-polyacrylamide gel electrophoresis, followed by autoradiography and quantitated by Phos-phorImager (Molecular Dynamics).

RESULTS
p38 MAPK Is Essential for T Cell Activation-T cell activation is accompanied by activation of p38 MAPK. Treatment of EL4 T cells with TPA/A23187 led to phosphorylation of p38 MAPK to an extent indistinguishable from stimulation with sorbitol or TNF-␣ (Fig. 1A). This was also confirmed by immunoprecipitation kinase assay using GST-ATF-2 (1-109) as substrate (Fig. 1B). Similar to anisomycin treatment, TPA/A23187 treatment significantly activated p38␣ in EL4 T cells. The p38 MAPK activation mediated by T cell activation was not limited to stimulation with TPA/A23187. Engagement of T cell receptor in EL4 cells by anti-CD3 antibody also induced activation of p38␣ (Fig. 1C). Activation of p38␣ was further enhanced when co-stimulated with anti-CD28. A similar extent of p38␣ activation by TCR engagement was found in T cell hybridomas 10I and 9C12.7 (52) as well as in the purified splenic T cells (Fig.  1C, not shown for 9C12.7 and splenic T cells). Both anisomycinand TCR-coupled p38␣ kinase activation was substantially inhibited by SB 203580 (10 M) (Fig. 1, B and C).
As report previously (37,38) and confirmed in our immunoblots (not shown), p38␣ and p38␦ are the major isoforms of p38 MAPK in T lymphocytes. p38␦ is not inhibited by SB 203580 FIG. 1. p38␣ but not p38␦ was activated by TCR engagement. A, EL4 T cells were activated by sorbitol (Sorb, 0.4 M), TNF-␣ (TNF, 100 ng/ml), or TPA (10 ng/ml) plus A23187 (80 ng/ml) (T/A) for 30 min, and cell extracts were prepared. The contents of phosphorylated p38 MAPK and p38 MAPK were determined by immunoblots with anti-phosphorylated T180/Y182 p38 MAPK antibody (New England Biolabs) and anti-p38␣ C-20 (Santa Cruz Biotech), respectively. B, EL4 T cells were treated with TPA/A23187 or with anisomycin (10 g/ml) in the absence or presence of SB 203580 (10 M). Cell lysates were prepared 20 min after activation, and 200 g of lysate was precipitated with 1 g of anti-p38␣ antibody (36) and 20 l of protein A-Sepharose. The kinase activity of the immune complexes was determined by the phosphorylation of GST-ATF-2 (1-109), and the reaction mixture was resolved on 15% SDS-polyacrylamide gel electrophoresis. C, EL4 and 10I T cells were stimulated with immobilized anti-CD3 (10 g/ml) or anti-CD3 (10 g/ml) plus anti-CD28 (2.5 g/ml) in the absence or presence of SB 203580 (10 M). Lysates were prepared 20 min after treatment, the kinase activity of p38␣ and p38␦ isolated by immunoprecipitation was determined with GST-ATF-2 (1-109) as substrate. D, EL4 T cells were stimulated with hydrogen peroxide (500 ng/ml) in the absence or presence of SB 203580 (20 M) for 20 min, and the kinase activity of p38␣ and p38␦ isolated by immunoprecipitation was measured using GST-ATF-2 as substrate. (35)(36)(37). We next examined whether p38␦ was activated by TCR engagement. Immunoprecipitation kinase assay indicated that, in contrast to p38␣, p38␦ was not activated by stimulation with anti-CD3 or anti-CD3 plus anti-CD28 (Fig. 1C). Neither did treatment of anisomycin activate p38␦ in T cells (not shown). On the contrary, hydrogen peroxide was an effective activator of both p38␣ and p38␦ in T cells (Fig. 1D). Hydrogen peroxide-activated p38␣, but not p38␦, was suppressed by SB 203580.
The induction of p38 MAPK activation was essential for TCR-mediated IL-2 production (46,48). We observed that IL-2 secretion was suppressed by 45% with the addition of SB 203580 at concentration as low as 0.625 M in concanavalin A-stimulated splenic T cells (not shown). Further inhibition was found at higher concentrations of SB 203580. The same extent of inhibition was seen in EL4 and 10I T cells (not shown).
Because the concentrations of SB 203580 (10 -20 M) used in the present study have been reported to inhibit JNK activation in monocytes and neuronal cells (59, 60), we tested whether JNK was similarly suppressed in T cells. As a control, TCRinduced ERK activation was not affected by SB 203580 (Fig. 2). Anti-CD3-induced JNK activation was slightly enhanced in the presence of SB 203580. Therefore, the effect observed with SB 203580 was not due to an inhibition of JNK and ERK in activated T cells.

Inhibition of p38 MAPK Prevented Activation-induced Cell Death but Had No Effect on Fas-initiated Apoptosis-Because
activation of p38 MAPK is essential for T cell activation, we investigated whether p38 MAPK was involved in activationinduced cell death. Activation of immature T cells such as T hybridomas by anti-CD3 induced cell death (52), as assessed by DNA fragmentation using fluorescence-activated cell sorter analysis (Fig. 3A). SB 203580 by itself did not trigger significant cell death during the time course of experiments (18 -24 h). However, as previously reported (61), prolonged incubation with SB 203580 (Ͼ24 h) did trigger apoptosis in T cells. All cell death analyses were thus conducted within 24 h period. Activation-induced cell death in T cell hybridoma 10I was suppressed by SB 203580, with a reduction in hypohaploid fraction from 45 to 15% (Fig. 3A). The extent of inhibition decreased with reduced concentrations of SB 203580 (Fig. 3B), yet antagonism on activation-induced death was evident with 5 M SB 203580. A similar inhibitory effect of SB 203580 was also observed in 9C12.7 and reactivated splenic T cells (not shown).
Previous reports have demonstrated that Fas-initiated apoptosis was completely resistant to SB 203580 (62,63). This is also confirmed by the fact that SB 203580 did not prevent apoptosis triggered by anti-Fas antibody (Jo2) in activated splenic T cells (Fig. 3C). Therefore, SB 203580 inhibited activation-induced cell death but not Fas-initiated cell death. The activation-induced death process that was antagonized by SB 203580 apparently must be at the stage prior to Fas-FasL ligand interaction. How TCR activation-induced Fas and FasL expression was modulated by SB 203580 was next examined.
p38 MAPK Was Essential for Activation-induced Fas/FasL Expression-Resting T cells express low level Fas. The activation by anti-CD3 triggered a significant increase of the cell surface Fas (Fig. 4A) and the Fas transcript (Fig. 4B). The expression of surface Fas on T cells was partially inhibited in the presence of 10 M SB 203580 (Fig. 4A, bold curve). A more prominent inhibition was observed with Fas mRNA (Fig. 4B), suggesting the requirement of p38 MAPK for Fas expression. SB 203580 did not completely suppress anti-CD3-induced Fas expression, in which both Fas mRNA and surface Fas expression remained abundant (Fig. 4).
Surface FasL protein and FasL mRNA was absent in resting T cells, and significant induction of surface FasL expression and FasL mRNA synthesis was observed followed TCR engagement (Fig. 4). In contrast to Fas, anti-CD3-triggered FasL mRNA increase was largely suppressed by SB 203580, suggesting that p38 MAPK may play a more critical role in FasL expression. Together, the suppression on activation-induced cell death by SB 203580 in 10I cells (Fig. 3) was due to an effective inhibition on FasL expression and a partial suppression on Fas expression. A preferential suppression on FasL expression by SB 203580 was also found in 9C12.7 cells (not shown).
Because inhibition of p38 MAPK only partially interfered with TCR-induced Fas expression, we tested whether the remaining Fas molecules still mediated apoptosis. Anti-Fas antibody Jo2 was unable to trigger apoptosis in resting T cells because the surface Fas expression was low (Fig. 3D). The Fas expression was up-regulated by stimulation with anti-CD3 for 12 h in the presence of SB 203580. The cells were then removed from stimulation and treated with Jo2 in the continued presence of SB 203580. Despite the fact that SB 203580 inhibited activation-induced cell death in T cells, the remaining Fas molecules on SB 203580-treated T cells were still functional as apoptotic-initiating molecules (Fig. 3D). Therefore, the inhibitory effect of SB 203580 on activation-induced death must be due to a preferential inhibition of FasL expression.
p38 MAPK Was Required for the FasL Promoter Activation-To further examine the induction of FasL, FasL promoter (Ϫ453 base pairs), which accounts for the inducibility (18,21), was isolated by PCR and was subcloned into luciferase reporter pGL2. The pGL2-FasL construct contained the transcription elements including NF-B, NF-AT, and AP-1. In this experiment, 9C12.7 cells were used because transfection efficiency of 10I cells was very low. Consistent with activation-induced FasL mRNA expression (Fig. 4), stimulation of 9C12.7 cells with anti-CD3 significantly activated the FasL promoter (Fig.  5A). As a control, no activation of pGL2 vector was detected in stimulated 9C12.7 cells. TCR-activated FasL promoter activity was largely inhibited by SB 203580 (Fig. 5A), suggesting that FasL promoter activation is dependent on p38 MAPK. There is a good correlation between the inhibition of FasL promoter activation and the suppression of FasL mRNA induction by SB 203580. Therefore, the inhibition on the FasL protein level and mRNA level by SB 203580 (Fig. 4) is partly attributed to a suppression on TCR-activated FasL promoter activity.
Activation of p38 MAPK Pathway Induced FasL Promoter Activation and FasL Expression in T Cells-The role of p38 MAPK was further elucidated by using the active forms of MKK3b and MKK6b (30), activators of p38 MAPK. Transfection of active MKK3b and MKK6b into EL4 cells led to activation of both p38␣ and p38␦ ( Fig. 6A; not shown for MKK6b). Overexpression of MKK3b or MKK6b activated the FasL promoter in EL4 T cells (Fig. 5B). Either MKK3b or MKK6b was less effective than TPA/A23187 in the activation of FasL promoter. For reasons that are unclear, MKK3b was a better activator of the FasL promoter than MKK6b (Fig. 5B).
In addition to the activation of FasL promoter, overexpression of active MKK3b resulted in a detectable FasL mRNA expression in EL4 T cells 24 h after transfection (Fig. 6B). In contrast, MKK3b alone was insufficient to increase the expression of Fas transcripts, suggesting the requirement of additional signals for Fas expression. Unlike TCR-coupled FasL expression (Fig. 4B), MKK3b-induced FasL was not reduced by SB 203580 treatment. Because p38␦ was resistant to SB 203580 inhibition (Fig. 6A), the difference in sensitivity to SB 203580 between TCR-induced and MKK3b-induced FasL expression was most likely due to the activation of p38␦ by MKK3b. This was supported by the suppression of MKK6-induced FasL expression through co-expression of the dominant negative form of p38␦ (Fig. 6C). The present results suggest that both p38␣ and p38␦ contribute to the expression of FasL, yet the selective activation of p38␣ by TCR confers the sensitivity of the FasL expression to SB 203580 inhibition.
FasL Expression Was Induced by Anisomycin in T Cells-We also examined whether the induction of FasL expression was limited to TCR engagement. p38 MAPK is activated by stress stimuli such as anisomycin (Fig. 7A). Anisomycin also stimulated FasL transcript expression in EL4 T cells 3 h after treatment (Fig. 7B). Anisomycin-induced FasL expression was partly suppressed by SB 203580. The sensitivity to SB 203580 may be explained by the observation that endogenous p38␦ was minimally activated by anisomycin in T cells. It may be noted that we failed to detect an induction of FasL expression in hydrogen peroxide-treated T cells because of the significant cell death early in the treatment.
FasL Expression Was p38 MAPK-dependent in 293T Cells-The involvement of p38 MAPK in FasL expression was not restricted to T cells. Treatment of 293T cells with anisomycin also led to activation of p38␣ (Fig. 8A) as well as expression of FasL (Fig. 8B). The effective inhibition of anisomycin-induced FasL expression by SB 203580 was correlated with a suppression of p38␣ activation. Therefore, p38 MAPK was also required for FasL induction in 293T cells. DISCUSSION p38 MAPK is activated by TCR engagement and TPA/ A23187 (43, 45-48) as well as by stress stimuli in T lympho-cytes. The stimulation of p38 MAPK by TCR was essential for IL-2 production (46,48). In this study, we demonstrated that TCR engagement activated p38␣ (Fig. 1C) and inhibition of p38␣ suppressed activation-induced cell death in T hybridoma (Fig. 3A). Our observation that p38 MAPK is essential for activation-induced T cell death is consistent with the observation that IgM-induced apoptosis of human B lymphocytes requires p38 MAPK (64) yet is in direct contrast to a recent report that antigen receptor-induced apoptosis is not affected by SB 203580 (45). The latter observation was made on the restimulation of lymph node T cells and on anti-IgM-induced WEHI 231 cells. We do not know the cause of such discrepancy because we found an inhibition of activation-induced apoptosis by SB 203580 on restimulated splenic T cells similar to that on T cell hybridomas.
Even though p38 MAPK is activated through Fas engagement, inhibition of p38 MAPK does not interfere with Fasmediated apoptosis (Fig. 3, C and D, and Refs. 45, 51, 63, and  65). Instead, the inhibition on activation-induced cell death was at the stage of activation-induced FasL and Fas expression (Fig. 4). Suppression by SB 203580 was greater on FasL ex- , and RNA was prepared 6 and 24 h after activation. 2 g of total RNA was used for cDNA synthesis by using oligo(dT) as primer. One-tenth of the cDNA synthesized was then amplified by using primers specific for Fas, FasL, and actin. 20% of PCR products were resolved on agarose gel for comparison.

FIG. 5. Activation of FasL promoter was p38 MAPK-dependent.
A, CD3-induced FasL promoter activation was suppressed by SB 203580. T cell hybridoma 9C12.7 was transfected with 5 g of pGL2-Basic or pGL2-FasL using the DEAE-dextran method. 1 g of pCH110 was included as an internal control. T cells were untreated or treated with anti-CD3 and/or SB 203580 after 24 h. Cell extracts were prepared after another 12 h, and the luciferase activity was determined. The luciferase activity was normalized by the ␤-galactosidase activity determined. B, activation of FasL promoter by MKK3b and MKK6b in EL4 cells. EL4 T cells were transfected with 4 g of pGL2-FasL together with 6 g each of an empty vector, active MKK3b, active MKK6b, or 3 g each of MKK3b and MKK6b. The luciferase activity was determined 24 h after transfection, except the cells stimulated with TPA/A23187 (24 h after transfection), and the luciferase activity was determined 24 h after activation.
p38 MAPK-dependent FasL Expression pression than on Fas expression. In addition, the residual surface Fas still mediated apoptosis in SB 203580-treated T cells (Fig. 3D), further supporting the idea that suppression on activation-induced apoptosis was due to the predominant inhibition of FasL expression. The prime role of p38 MAPK in the induction of FasL is also illustrated by the overexpression of MKK3b leading to expression of FasL in T cells (Fig. 6B). In contrast, despite the participation of p38 MAPK in Fas expression (Fig. 4), activation of p38 MAPK alone did not increase Fas expression (Fig. 6B).
p38 MAPK-dependent FasL expression is attributed to a requirement for p38 MAPK for FasL promoter activation, as demonstrated by the sensitivity of TCR-induced FasL promoter activation to SB 203580 (Fig. 5A) and by the induction of FasL promoter by MKK3b and MKK6b (Fig. 5B). The transcriptional elements necessary for FasL expression, including AP-1, NF-B, and NF-AT, have recently been identified (12,(15)(16)(17)(18)(19)(20)(21). TCR-induced NF-AT activation is inhibited by SB 203580 (45). The contribution of p38 MAPK in TNF-␣-induced NF-B activation has been documented (66,67). On a preliminary study, we also observed that the activation of the analogous NF-B and NF-AT elements were partially inhibited in presence of SB 203580. 2 Analysis of the NF-AT and B elements from FasL promoter are currently being conducted. It is likely that the inhibition of FasL expression by SB 203580 may be attributed to a specific inhibition of NF-AT and NF-B.
We have also observed a preferential activation of p38␣, but not p38␦, by TCR engagement (Fig. 1C). In addition, p38␦ was stimulated by hydrogen peroxide but not by anisomycin in T cells. Therefore, similar to that previously reported (36,37), p38␣ and p38␦ are differentially regulated in T cells. Interestingly, the induction of FasL was no longer sensitive to SB 203580 when both p38␣ and p38␦ were activated by MKK3b (Fig. 6B). This is consistent with the known resistance of p38␦ to SB 203580 (35)(36)(37). A role of p38␦ in FasL expression was supported by the inhibition of FasL expression through the co-transfection of p38␦ dominant negative mutant (Fig. 6C). Hence both p38␣ and p38␦ contribute to FasL expression, and the sensitivity of TCR-induced FasL expression to SB 203580 is a consequence of the selective activation of p38␣ in T cells. Our results may serve as another example that the use of pyridinyl imidazole inhibitor is highly dependent on the isoforms of the p38 MAPK that are activated. p38 MAPK is known to be activated by stress stimuli such as TNF-␣, IL-1, UV, and ␥-irradiation. Interestingly, Fas and FasL are up-regulated by stress stimulation including ␥-irradiation, UV irradiation, and cytotoxic drugs (6 -12). In this study, we also demonstrate that activation of p38 MAPK by anisomycin induced FasL expression (Fig. 7). This result suggests that, in addition to TCR-coupled signaling, p38 MAPK also mediates the induction of FasL by stress signals.
It may be noted that the present result does not imply that p38 MAPK alone accounts for FasL expression under physiological and pharmacological stimulation. Stimulation with anti-CD3 or anisomycin activated kinases other than p38 MAPK. The activation of FasL expression by p38 MAPK was performed through overexpression of MKK3b/MKK6b at levels that were likely unphysiological. In addition, MKK3/6 activates signaling pathways that are p38 MAPK-independent (62,63). Therefore, the present study cannot be used to argue against the requirement of other signal pathways for FasL expression. Furthermore, overexpression of MEKK1 has been shown to induce JNK-dependent FasL expression (11,15,60). A recent study also indicates that ERK is required for activation-induced FasL expression (68). A full activation of FasL expression under physiological condition likely requires the coordination of p38 MAPK with other activation signals. We are currently investigating the integration of different activation signals in the induction of FasL.
Recent studies suggest that p38 MAPK is involved in a number of apoptotic processes. Activation of p38 MAPK is critical for apoptosis induced by nerve growth factor deprivation in PC12 cells (69). p38 MAPK mediates apoptosis induced by withdrawal of insulin in primary neuron culture (70). Inhibition of p38 MAPK activation prevents glutamate-induced apoptosis in rat cerebellar granule cells (71). Opposite effect on apoptosis between p38␣ and p38␤ have also been found (61,72). Results from the present study suggest that p38 MAPK mediates FasL expression. It will be interesting to examine whether the above-mentioned apoptosis involves Fas-FasL interaction. The increased expression of FasL mediated by MKK3/6 may also explain why MKK3/6 activation enhances Fas-mediated apoptosis (62).